Convergent moral prohibitions
A Socratic walk-through of convergent moral prohibitions — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why do societies that never met converge on forbidding the same short list of acts?
Cultures differ about nearly everything a moral code can cover — marriage, property, food, revenge, what is owed to strangers, what may be done to the dead. Against that background the agreements are conspicuous. Killing a fellow member of your own community, taking their things, breaking your word to them, and betraying those who depend on you turn up as prohibited in societies with no contact and no common ancestry within recorded memory.
The tempting reading is that each of them saw the same truth. But that explains a convergence by crediting everyone with the same insight, which is a way of restating the puzzle rather than solving it. There is a different question available, and it is the one that usually pays: not why so many groups arrived at this answer, but what became of the groups that tried the alternatives.
Reasoning it through
REASONING #Before asking what survived, get the invariant right, because the usual statement of it is far too generous.
Did no society permit killing? Plainly many did — in feud, in raiding, in the exposure of infants, in sacrifice, in punishment, and in war on outsiders. Did none permit taking? Raiding a neighbouring group's herds has been honourable in a great many places. So the shared prohibition is not "do not kill" or "do not take". It is something much narrower: do not do these things to members of the cooperating unit, and it comes bundled with an enormously variable answer to who that unit contains. Notice that this is already a clue. A filter operating on whether groups hold together would produce exactly this shape — rules that protect the unit, indexed to the unit, silent about everyone else.
What is the state of the evidence? The most systematic recent attempt is Oliver Scott Curry, Daniel Mullins and Harvey Whitehouse's 2019 study in Current Anthropology, which coded ethnographic records from sixty societies and reported seven cooperative behaviours — helping kin, helping your group, reciprocating, being brave, deferring to superiors, dividing disputed resources, and respecting prior possession — appearing as morally good wherever they appeared, with no clear counter-examples. That is real evidence and it is also contested: coding ethnography for moral content involves judgement calls, and the sample is not a random draw from human history. Take it as strong support for a thin universal list, not as a settled census.
Now the selection question. Three explanations are on offer, and they are not mutually exclusive. The first is diffusion — they all learned it from somewhere. That one can be tested and largely fails: populations isolated for millennia, such as Tasmania or the pre-contact New Guinea highlands, carry the same core prohibitions, so the convergence cannot be borrowed. The second is a shared evolved psychology: we are all running similar machinery for detecting cheats, caring for kin, and resenting harm, so we keep generating similar rules. The third is a filter on the groups themselves — cultural group selection, in Boyd and Richerson's formulation — in which communities whose norms sustain cooperation grow, absorb, or outlast those whose norms do not.
The third is the one worth pressing on, because it does not require anyone to be right about anything. Imagine a society that genuinely permits its members to kill and rob each other at will. What happens to it? It is not merely unpleasant; it cannot run. Nobody can store grain, plan a harvest, raise children jointly, or leave the settlement. And the crucial point for our purposes is that such a group cannot transmit its norms either, because transmission requires the very continuity the norms destroy. The alternatives are not out-argued. They are unstable, and instability removes them from the record without anyone noticing.
That is convergence by elimination rather than by inspiration, and it explains something the inspiration story cannot: why the shared list is short, thin, and structural, rather than rich and specific. Only the rules that are load-bearing for the group's continued existence get filtered for. Everything else — diet, dress, sexual arrangements, obligations to outsiders — is free to vary, and does.
The analogy
THE ANALOGY #Every seagoing culture converged on a hull that tapers at the bow, and none of them copied the others. It was not shared genius. Water rejects the alternatives, and it rejects them quickly, so the surviving design space is narrow no matter who is doing the designing.
a bad boat sinks on its first voyage, visibly and at once, whereas a society can carry a terrible norm for centuries — chattel slavery, foot-binding, judicial torture — so the moral filter is slow, weak, and noisy, and nothing entitles us to read "widespread" as "passed a test".
Clarifying the model
THE MODEL #Two corrections, one of which matters a great deal. The small one: this is not an argument that the convergent rules are innate. A filter on which norm-sets persist would produce the same distribution whether the rules were reinvented, taught, or felt, and disentangling the three is exactly what the current research is trying to do.
The large one: an explanation of why a rule is widespread is not a demonstration that it is right, and the argument runs in the wrong direction if you try to use it that way. Selection favours what keeps a group going, which is not the same as what is just — it is entirely capable of favouring rigid hierarchies, harsh treatment of outsiders, and the suppression of dissent, and historically it plainly has. The convergence is a fact about the distribution of norms, and it leaves every question about their justification exactly where it was.
A picture of it
THE PICTURE #How to readthe three left-hand branches are rival explanations of the same observation, of which only the first has been largely excluded — and the fourth branch is the reminder that what converged is thin, because the boundary of the protected group did not converge at all.
What became clearer
WHAT CLEARED #Asking what happened to the alternatives, rather than what everyone discovered, changes the shape of the answer. The short prohibited list looks less like a truth repeatedly glimpsed and more like the residue of a filter: norm-sets that permit their members to prey on each other cannot hold a group together long enough to be passed on, so they leave no descendants and no record. That accounts for the convergence, for its thinness, and for its sharp restriction to insiders — and it accounts for none of its authority.
Where to go next
ONWARD #- Whether cultural group selection is a real mechanism or a redescription: the ongoing dispute about groups as units of selection.
- Donald Brown's Human Universals (1991) and how its list compares with the cooperation findings.
- The variable boundary: what determines how wide the protected circle is drawn, and why it has widened historically.
- The is/ought problem in its evolutionary form — what, if anything, convergence could ever contribute to justification.
Key terms
TERMS #| Term | What it means |
|---|---|
| Cultural group selection | the proposal that competition and differential persistence among groups, not just among individuals, shapes which norms survive. |
| Moral universals | the short list of evaluations that appear across the ethnographic record, as distinct from specific rules, which vary widely. |
| In-group indexing | the fact that convergent prohibitions protect members of the cooperating unit while leaving the unit's boundary undefined and highly variable. |
Every term the collection defines is gathered in the glossary.